On Thu, May 03, 2012 at 04:37:49PM -0700, Andrew Morton wrote:

On Thu,  3 May 2012 22:45:12 +0530
rajman mekaco [off-list ref] wrote:

quoted

When CONFIG_HUGETLB_PAGE_SIZE_VARIABLE is not defined, then
pageblock_default_order has an argument to it.

However, free_area_init_core will call it without any argument
anyway.

Remove the argument to pageblock_default_order when
CONFIG_HUGETLB_PAGE_SIZE_VARIABLE is not defined.

Signed-off-by: rajman mekaco <redacted>
---
 mm/page_alloc.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index a712fb9..4b95412 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c

@@ -4274,7 +4274,7 @@ static inline void __init set_pageblock_order(unsigned int order)
  * at compile-time. See include/linux/pageblock-flags.h for the values of
  * pageblock_order based on the kernel config
  */
-static inline int pageblock_default_order(unsigned int order)
+static inline int pageblock_default_order(void)
 {
 	return MAX_ORDER-1;
 }

Interesting.  It has been that way since at least 3.1.

/me slaps self

It didn't break the build because pageblock_default_order() is only
ever invoked by set_pageblock_order(), with:

	set_pageblock_order(pageblock_default_order());

and set_pageblock_order() is a macro:

#define set_pageblock_order(x)	do {} while (0)

There's yet another reason not to use macros, dammit - they hide bugs.


Mel, can you have a think about this please?  Can we just kill off
pageblock_default_order() and fold its guts into
set_pageblock_order(void)?  Only ia64 and powerpc can define
CONFIG_HUGETLB_PAGE_SIZE_VARIABLE.

This looks reasonable to me.

-- 
Mel Gorman
SUSE Labs